INDEX

Explanations

words related to specific entities with high relevance

oai_token-act-pair · gpt-3.5-turbo

proper nouns or specific names and entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 millenn

-0.83

 carbohyd

-0.82

 referen

-0.82

ccording

-0.71

 tremend

-0.68

 sympathetic

-0.67

 encount

-0.67

Þ

-0.66

 looph

-0.66

 forc

-0.65

POSITIVE LOGITS

phia

0.98

Â®

0.94

 Labs

0.87

idae

0.86

 Chronicles

0.84

 Mountains

0.83

 Productions

0.83

 Valley

0.82

 Studios

0.81

 Records

0.79

Activations Density 0.385%

words related to specific entities with high relevance

proper nouns or specific names and entities

No Comments

No Known Activations